Tight-Binding Inhibition of Human Monoamine Oxidase B by

Apr 12, 2018 - ABSTRACT: Monoamine oxidase B (MAO-B) is a validated drug target for Parkinson's disease. Chromone derivatives were identified as novel...
0 downloads 0 Views 4MB Size
Download PDF
Article pubs.acs.org/jmc

Cite This: J. Med. Chem. 2018, 61, 4203−4212

Tight-Binding Inhibition of Human Monoamine Oxidase B by Chromone Analogs: A Kinetic, Crystallographic, and Biological Analysis Joana Reis,† Nicola Manzella,‡ Fernando Cagide,† Jeanne Mialet-Perez,‡ Eugenio Uriarte,§,∥ Angelo Parini,‡ Fernanda Borges,† and Claudia Binda*,⊥ †

CIQUP/Department of Chemistry and Biochemistry, University of Porto, 4169-007 Porto, Portugal Institute of Metabolic and Cardiovascular Diseases (I2MC), Institut National de la Santé et de la Recherche Médicale (INSERM), Université de Toulouse, 31432 Toulouse Cedex 4, France § Departamento de Química Orgánica, Facultad de Farmacia, Universidad de Santiago de Compostela, 15782 Santiago de Compostela, Spain ∥ Applied Chemical Science Institute, Universidad Autonoma de Chile, 7500912 Santiago de Chile, Chile ⊥ Department of Biology and Biotechnology, University of Pavia, 27100 Pavia, Italy ‡

S Supporting Information *

ABSTRACT: Monoamine oxidase B (MAO-B) is a validated drug target for Parkinson’s disease. Chromone derivatives were identified as novel potent and reversible MAO-B inhibitors, and herewith we report on a crystallographic and biochemical analysis to investigate their inhibition mechanism. The crystal structures of human MAOB in complex with three chromone analogs bearing different substituents on the exocyclic aromatic ring (determined at 1.6−1.8 Å resolution) showed that they all bind in the active site cavity of the protein with the chromone moiety located in front of the FAD cofactor. These inhibitors form two hydrogen bonds with Tyr435 and Cys172 and perfectly fit the hydrophobic flat active site of human MAO-B. This is reflected in their tight-binding mechanism of inhibition with Ki values of 55, 17, and 31 nM for N-(3′,4′dimethylphenyl)-4-oxo-4H-chromene-3-carboxamide (1), N-(3′chlorophenyl)-4-oxo-4H-chromene-3-carboxamide (2), and N-(3′-fluorophenyl)-4-oxo-4H-chromene-3-carboxamide (3), respectively. These compounds were also 1000-fold more effective than L-deprenyl in reducing the cellular levels of reactive oxygen species (ROS).

■

INTRODUCTION Monoamine oxidases (MAOs, EC 1.4.3.4) are broadly distributed enzymes that contain a flavin adenine dinucleotide (FAD) cofactor covalently bound to a cysteine residue. They are expressed in several living organisms, and two isoforms are present in mammals, namely, MAO-A and MAO-B, which are responsible for the major neurotransmitter metabolism in both the central nervous system (CNS) and peripheral tissues.1,2 Both isoenzymes are bound to the outer mitochondrial membrane and catalyze the oxidative deamination of aromatic amine substrates (Figure 1). MAO-A and MAO-B display different substrate specificity that depends on the tissue distribution of these enzymes. Serotonin and noradrenaline are preferentially metabolized by MAO-A, whereas dopamine and phenethylamine are mainly metabolized by MAO-B. In light of their key role in neurotransmitter metabolism in the brain, these enzymes are validated drug targets for neurological diseases. MAO inhibitors © 2018 American Chemical Society

Figure 1. Scheme of the FAD-dependent oxidative reaction catalyzed by MAOs. Dopamine is shown as substrate that is preferentially metabolized by MAO-B in the CNS. Besides the aldehyde product (dopanal) and ammonia derived from substrate oxidation, hydrogen peroxide is also generated upon cofactor reoxidation by molecular oxygen.

are currently used in the clinical practice for the treatment of depression and Parkinson’s disease. Most of them, such as Received: March 6, 2018 Published: April 12, 2018 4203

DOI: 10.1021/acs.jmedchem.8b00357 J. Med. Chem. 2018, 61, 4203−4212

Journal of Medicinal Chemistry

Article

Figure 2. Chemical structures of MAO-B inhibitors based on the chromone scaffold. Safinamide is also shown because a comparative analysis with this clinically used inhibitor is carried out in the text.

the recent findings highlighting the role of MAO-derived ROS in several cellular processes, we have also evaluated the biological activity of these compounds in reducing the cellular ROS levels in HEK-293 cells.

tranylcypromine and rasagiline, irreversibly inactivate the enzyme by forming a covalent adduct with the flavin cofactor. More recently, safinamide, a MAO-B selective and reversible inhibitor, was newly approved in Europe and United States as Xadago as adjunct therapy for Parkinson’s disease.3 Additionally, the expression of MAO-B in neuronal tissues increases 4fold with aging, resulting in an intensification of dopamine metabolism and higher production of hydrogen peroxide.4 Indeed, in the clinical treatment of Parkinson’s disease MAO-B inhibition is thought to be beneficial both in restoring physiological levels of dopamine and in exerting a neuroprotective effect by preventing the production of excessive hydrogen peroxide (Figure 1).4 This aspect is emerging as a relevant issue also in the context of MAO activity and reactive oxygen species (ROS) in non-neuronal tissues including heart,5 prostate,6 and adipose macrophages.7 Chromones (4H-benzopyran-4-one), a class of heterocyclic compounds largely distributed in nature, have attracted attention in the field of drug discovery due to their interesting biological activities (Figure 2).8,9 Moreover, several studies have shown that chromone is a privileged structure for the rational discovery and development of new MAO-B inhibitors.10−13 Inhibition studies performed with microsomal recombinant human MAO-B showed that a series of chromone analogs selectively and reversibly bind to this enzyme with IC50 values in subnanomolar range (1 and 2, Figure 2).13 Preliminary kinetic studies demonstrated that compound 1 acted as a competitive MAO-B inhibitor while compound 2 appear to display a noncompetitive profile.13 Despite several computational docking studies to predict chromone binding to MAOs,10,13,14 no experimental structural information was available so far. In order to clarify the MAO-B inhibition mechanism of this class of compounds, we undertook a crystallographic analysis combined with a more in-depth kinetic characterization using purified recombinant human MAO-B. The study was performed with chromones 1−3 to probe the effect of different substituents on the aromatic ring and with an analog of 2 lacking the carbonyl group in the linker moiety (chromone 4, Figure 2) to investigate the role of hydrogen bonding (see later in the text). Moreover, in light of

■

RESULTS AND DISCUSSION Chemistry. Compounds 1 and 2 were previously synthesized (Scheme 1) in moderate yields through the in

Scheme 1. Strategy for the Synthesis of Chromone-3phenylcarboxamide Derivatives 1−3 from Chromone 3Carboxylic Acida

a

Reagents and conditions: (a) aniline derivatives, POCl3, DMF, rt, 1− 5 h.

situ generation of an acyl chloride intermediate, using phosphoryl chloride (POCl3) in N,N-dimethylformamide (DMF) and the subsequent addition of the appropriate arylamine, as reported by Reis et al.13 Likewise, compound 3 was synthesized from chromone 3-carboxylic acid and 3fluorobenzylamine in an amidation reaction using POCl3 (Scheme 1). Compound 4 was obtained by a reductive amination, in the presence of sodium triacetoxyborohydride, of the imine intermediate formed from the reaction of 3formylchromone with 3-chloroaniline (Scheme 2). Structures of Human MAO-B in Complex with Chromone Inhibitors. The crystal structures of human 4204

DOI: 10.1021/acs.jmedchem.8b00357 J. Med. Chem. 2018, 61, 4203−4212

Journal of Medicinal Chemistry

Article

The pyrone carbonyl group points toward the upper part of the active site, whereas the oxygen atom of the ring is located in the more hydrophilic part of the cavity and establishes a weak hydrogen bond with Tyr435 (about 3.4 Å in all structures). This binding mode is in agreement with previous docking studies11,14 except for a hydrogen bond predicted to exist between the pyrone carbonyl group and Tyr326 that instead is not observed. The amide linker of all three inhibitors lies near the gating residue Ile199 that is in the open conformation similarly to other human MAO-B structures in complex with bulky inhibitors.1,15 The carbonyl group of the linker is H-bonded to Cys172 with distances ranging from 3.2 to 3.4 Å (Figure 3b− d), which was also predicted by previous docking studies and which is fully consistent with the longer hydrogen bonds that are known to involve sulfur atoms.16 Interestingly, careful inspection of the electron density showed that the side chain of Cys172 exists in a double conformation (Figure 3b−d). Accordingly, the program PHENIX17 was used to refine the occupancy values for the two rotamers of Cys172 in each of the three structures. This resulted in 0.5 in the case of the complex with inhibitor 1, whereas for the structures with 2 and 3 the conformation H-bonded with the inhibitor was preferred with respect to the other one (occupancy is about 0.7 versus 0.3 of the other rotamer; Figure 3c,d). Interestingly, the cysteine rotamer that is H-bonded with the chromone linker moiety is the conformation normally present in other human MAO-B structures (Figure 4). Nevertheless, Cys172 double conformation was previously observed in the structures of human MAOB in complex with the slow substrate p-NO2-benzylamine18 and in complex with a rasagiline analog bearing an OH substituent on the aminoindan ring.18,19 In both cases, one of the two

Scheme 2. Strategy for the Synthesis of Chromone Derivative 4 from 3-Formylchromonea

a Reagents and conditions: (a) 3-chloroaniline, Na(AcO)3BH, rt, 15 min.

MAO-B in complex with inhibitors 1−3 were solved at 1.8, 1.6, and 1.7 Å resolution, respectively. Crystallographic statistics are reported in Table 1. The unbiased electron density map was of extraordinary good quality, which allowed unambiguous modeling of the inhibitor molecules in the enzyme active site. For all three structures, no relevant differences in inhibitor binding and active site conformation were found between the two molecules present in the asymmetric unit (rmsd for the Cα atoms are 0.28, 0.23, and 0.22 Å between monomers A and B for the structures in complex with compounds 1, 2, and 3, respectively). Hereafter we will refer to subunit A of each structure for the following discussion. All chromone analogs clearly bind in the hydrophobic active site of human MAO-B1 (Figure 3a). In particular, the 1,4benzopyrone moiety occupies the substrate-binding site in front of the flavin ring, whereas the amide linker and the differently substituted exocyclic ring moieties extend into the entrance space whose access is negotiated by loop 99−105 (Figure 3).

Table 1. Data Collection and Refinement Statistics for the Crystal Structures of Human MAO B in Complex with Chromone Inhibitors chromone 1 space group unit cell axes (Å)

resolution (Å) Rsyma,b (%) CC1/2 (%)b completenessb (%) unique reflections redundancyb I/σb no. of non-hydrogen atoms protein/FAD/inhibitor detergentc/glycerol/water average B value for protein/inhibitor atoms (Å2) b,d Rcryst (%) Rfreeb,d (%) rms bond length (Å) rms bond angles (deg)

chromone 2

chromone 3

C222 a = 131.2 b = 222.7 c = 86.5 1.8 16.4 (97.1) 99.4 (53.5) 100.0 (100.0) 117,061 6.8 (6.4) 10.2 (1.5)

C222 a = 131.4 b = 222.0 c = 86.1 1.6 11.4 (99.4) 99.6 (45.9) 99.9 (100.0) 164,973 4.5 (4.3) 8.6 (1.3)

C222 a = 131.7 b = 222.0 c = 86.3 1.7 10.5 (66.2) 99.5 (66.2) 99.4 (98.9) 137,691 4.5 (4.4) 10.1 (1.9)

7943/2×53/2×22 26/2×6/846

7943/2×53/2×21 26/2×6/1051

7933/2×53/2×21 27/2×6/1012

17.9/20.2 16.8 (27.1) 19.9 (29.3) 0.011 1.48

17.9/19.4 16.5 (27.9) 19.1 (27.3) 0.009 1.42

16.6/17.1 16.2 (24.7) 18.9 (28.9) 0.010 1.47

Rsym = ∑|Ii − ⟨I⟩|/∑Ii, where Ii is the intensity of ith observation and ⟨I⟩ is the mean intensity of the reflection. bValues in parentheses are for reflections in the highest resolution shell. cAs in previous human MAO-B structures, one molecule of the Zwittergent 3-12 detergent (used in crystallization experiments) is partly visible in the electron density of each of the two protein monomers present in the asymmetric unit (15 and 11 atoms out of 22 in monomers A and B, respectively, except for complex with chromone 3 where in monomer A 16 detergent atoms out of 22 are visible). dRcryst = ∑|Fobs − Fcalc|/∑|Fobs| where Fobs and Fcalc are the observed and calculated structure factor amplitudes, respectively. Rcryst and Rfree were calculated using the working and test sets, respectively. a

4205

DOI: 10.1021/acs.jmedchem.8b00357 J. Med. Chem. 2018, 61, 4203−4212

Journal of Medicinal Chemistry

Article

Figure 3. Crystal structure of human MAO-B in complex with chromone inhibitors. (a) Overall ribbon diagram (ice blue) of human MAO-B structure oriented with the membrane-bound C-terminal helix pointing at the bottom of the figure. The loop formed by residues 99−105 that admits to the enzyme active site is highlighted in bold style. The FAD cofactor is in stick representation with carbon, nitrogen, oxygen, and phosphorus atoms colored in yellow, blue, red, and magenta, respectively. The structure in complex with inhibitor 1 (PDB code 6FVZ) is shown as reference of chromone analogs binding mode with atom color code as FAD except for carbons that are in dark green. (b) Zoomed view of human MAO-B active site in complex with inhibitor 1 (PDB code 6FVZ). The orientation of the molecule is rotated approximately 45° around an axis perpendicular to the figure plane with respect to Figure 3a. Residues lining the enzyme active site are depicted as sticks with carbon atoms in ice blue (sulfur atoms in green). Hydrogen bonds are drawn as dashed lines. Water molecules are represented as red spheres. The refined 2Fo − Fc electron density map contoured at 1.2σ is shown in dark red for the inhibitor molecule and for Cys172 in double conformation. (c) Zoomed view of human MAO-B active site in complex with inhibitor 2 (PDB code 6FW0). Orientation and color code is as in Figure 3b (chlorine atom in cyan). (d) Zoomed view of human MAO-B active site in complex with inhibitor 3 (PDB code 6FWC). Orientation and color code is as in Figure 3b (fluorine atom in pink).

The differently substituted aromatic ring linked to the chromone unit through the amide spacer represents the varying moiety of the inhibitors under study (Figure 2). In all structures the substituted exocyclic ring binds in the entrance space of MAO-B active site with the halogen atom of 2 and 3 occupying the same position as one of the two methyl groups of compound 1 (Figure 4). No differences in the active site architecture were found between 2 and 3, indicating that the halogen atomic radius does not affect inhibitor binding.

Cys172 rotamers forms a hydrogen bond with the oxygen atom of the ligand, similar to that formed with the carbonyl group of the chromone inhibitors. These observations suggest that the involvement of Cys172 in hydrogen bond formation as well as its flexibility in adopting a double conformation, though rarely found in other MAO-B crystal structures,1 is not exclusive to the complexes with chromones and might be relevant to stabilize binding of inhibitors bearing an oxygen atom in proximity of Cys172. 4206

DOI: 10.1021/acs.jmedchem.8b00357 J. Med. Chem. 2018, 61, 4203−4212

Journal of Medicinal Chemistry

Article

3 (Figure 4) suggesting that this part of the active site may represent a subsite favorable for halogen binding. This comparative analysis also highlighted another aspect related to loop 99−105 whose flexibility allows the admission of ligands into the enzyme active site (Figure 3a).1 In particular, Pro102 and Phe103 were previously found to adopt different conformations depending on the size of the bound inhibitor, as observed also in the structures in complex with chromone analogs described in this study (Figure 4). Nevertheless, in the case of compounds 2 and 3, we observed a shift of about 1.5 Å in the position of Pro104-Pro105 toward the inhibitor, which was not found in any other MAO-B structure including that in complex with 1. This conformation is probably due to the fact that the halogenated chromone analogs do not extend as long as safinamide and lack the second substituent on the aromatic ring as it happens in 1 (Figure 4). However, as this conformational difference is moderate and does not correlate with binding affinity of chromone analogs, this aspect is not worth further discussion. An important feature in the conformation of all chromone analogs bound to human MAO-B active site is the intramolecular hydrogen bond between the carbonyl group of the pyrone ring and the nitrogen atom of the spacer (about 2.7 Å in all three structures; not shown in Figure 3b−d). This interaction, which was already found in chromone compounds as demonstrated by single-crystal X-ray diffraction,12 lends a planar and rigid conformation to the molecule. The human MAO-B active site cavity, with its flat and elongated shape,1

Figure 4. Structural superposition between structures of human MAOB in complex with inhibitor 1, inhibitor 2, and safinamide15 (PDB code 2V5Z) depicted with carbon atoms in gray, white, and black, respectively (inhibitor 3, whose fluorine-substituted aromatic ring perfectly overlaps with that of 2, was omitted for clarity). Color code for FAD and the other atoms is as in Figure 3.

Interestingly, structure superposition with MAO-B in complex with safinamide (Figure 4) and coumarin analogs15 showed that their halogen substituents perfectly overlaps with those of 2 and

Figure 5. Kinetic studies on the mechanism of human MAO-B inhibition by chromones 1 (a) and 2 (b). The effect of the inhibitors on the enzyme was determined using both the nonlinear regression on GraphPad Prism software 5.0 (left panels) and the double reciprocal plot analysis (right panels). Both apparent kcat (y axis) parameters were calculated as v0/[MAO-B]. 4207

DOI: 10.1021/acs.jmedchem.8b00357 J. Med. Chem. 2018, 61, 4203−4212

Journal of Medicinal Chemistry

Article

provides an ideal niche for this class of molecules, which may be related to their inhibition mechanism as described below. Inhibition Studies on Chromone Analogs. The crystallographic analysis demonstrated that all chromone inhibitors bind in the MAO-B active site in a highly conserved way, which appears in contrast with previous findings suggesting that 1 and 2 may target different sites on the enzyme on the basis of preliminary inhibition assays.13 To clarify the inhibition mechanism of this class of compounds, we carried out a more in-depth biochemical characterization using a fluorimetric horseradish peroxidase-Amplex Red-coupled assay20 with a purified sample of recombinant human MAOB and benzylamine as substrate. The high sensitivity of this assay allows the measurement of enzyme activity in steady-state conditions even with nanomolar affinity ligands (i.e., MAO-B 0.7 nM final concentration). Initial velocity data measured at varying substrate and inhibitor concentrations were fitted to the appropriate form of the Michaelis−Menten equation (using GraphPad Prism software 5.0) which gave the best graphical fitting to the curves with a R2 value proximal to 1.0. Moreover, in order to gain a better visual assessment of enzyme inhibition, data were also analyzed using a linear double-reciprocal plot. In the case of inhibitor 1, the nonlinear regression analysis gave the best fitting with the equation corresponding to a mixed type inhibition mechanism with Ki = 0.31 ± 0.08 nM (α = 25.90) (Figure 5a, left panel). Indeed, in the double-reciprocal plot data were fitted to lines intersecting in the upper-left quadrant (Figure 5a, right panel), indicating that this compound acted as a mixed-type inhibitor toward human MAO-B. Instead, chromone 2 clearly showed a noncompetitive inhibition profile (Ki = 7.5 ± 1.2 nM) by both Michaelis−Menten equation fitting and linear regression plot analysis (Figure 5b). These data are in agreement with the previous studies performed with the microsomal enzyme and reveal that, independent from the substituent on the aromatic ring, these chromone analogs seemingly fail to feature a competitive mechanism of inhibition. By use of the same fluorimetric assay, all compounds were also tested on purified human MAO-A with kynuramine as substrate and, as previously found,13 no enzyme inhibition was detected up to 100 μM. It is well-known that this mixed or noncompetitive behavior might occur when ligands bind to the enzyme active site in a way that is defined as “tight-binding inhibition”.21 In this case the standard steady-state kinetic model used to describe the mechanism of inhibition is no longer valid and the resulting double-reciprocal plots appear similar to the classical pattern for noncompetitive inhibitors. To test whether this may hold for the chromone analogs, we applied a protocol by plotting the IC50 values for each chromone analog determined at different substrate concentrations,21 using the same fluorimetric assay employed for the classical steady-state experiments. In the case of tight-binding ligands, higher concentrations of substrate are needed in order to compete with the inhibitor in binding to the enzyme active site. With both compounds 1 and 2 (Figure 6), a linear correlation between IC50 values and substrate concentration was observed, in full agreement with tight-binding competitive inhibition.21 After having assessed that chromone analogs feature a tightbinding inhibition mechanism, the next step was to determine an accurate value of inhibition constants. This was accomplished by following published protocols,21,22 which use the Morrison equation (eq 1), to describe the fractional velocity of

Figure 6. Effects of substrate (S) concentration on the IC50 values for chromone 2. A similar pattern was obtained for chromone 1 (data not shown).

an enzymatic reaction as a function of inhibitor concentration, at fixed concentrations of enzyme and substrate: ([E] + [I] + K iapp) − vi =1− v0

([E] + [I] + K iapp)2 − 4[E][I] 2[E]

(1)

where vi/v0 represents the enzyme activity, whereas [E] is the total enzyme concentration, [I] is the total inhibitor concentration, and Kapp is the apparent inhibition constant. i The real inhibition constant (Ki) is obtained using a different equation in accordance with the inhibitor type. In the case of a competitive tight-binding inhibitor, this is achieved through eq 2: ⎛ [S] ⎞ K iapp = K i⎜1 + ⎟ Km ⎠ ⎝

(2)

Accordingly, the fluorimetric Amplex Red assay was used to obtain the data that were plotted in a dose−response curve described by the Morrison equation for both chromone 1 (Figure 7a) and chromone 2 (Figure 7b). By use of eq 2, the tight-binding Ki values for the inhibitors were determined to be 55 ± 9 nM and 17 ± 2 nM for 1 and 2, respectively. On the basis of these results, we extended the kinetic analysis to other chromone analogs to probe both the type of halogen substituent on the exocyclic phenyl ring and the importance of the carbonyl group on the amide spacer (forming the hydrogen bond with Cys172) in enhancing binding affinity. For the former task, inhibitor 3 bearing a fluorine group (Figure 2) and its bromine analog N-(3-bromophenyl)-4-oxo-4H-chromene-3carboxamide13 were tested for their binding mechanism using the procedures described above for 1 and 2. Both 3 and the bromine analog showed a tight-binding profile with the nonlinear regression curves and double-reciprocal plots displaying a mixed and a noncompetitive type of inhibition, respectively (data not shown). Accordingly, the tight-binding inhibition Ki values were determined by dose−response curve analysis (Figure 8a,b), which resulted in 31 ± 2 nM for compound 3 and 27 ± 2 nM for N-(3-bromophenyl)-4-oxo4H-chromene-3-carboxamide. This indicates that all chromone analogs bearing a different halogen substituent in meta position of the aromatic ring bind to human MAO-B with similar affinity, which nevertheless was found also when two methyl groups are present (chromone 1). To investigate the role of the amide spacer and its interaction with Cys172, a chromone analog bearing a chlorine group in meta position but lacking the carbonyl group in the linker moiety (chromone 4, Figure 2) was synthesized and used for both kinetic and crystallographic analysis. Even though a tight4208

DOI: 10.1021/acs.jmedchem.8b00357 J. Med. Chem. 2018, 61, 4203−4212

Journal of Medicinal Chemistry

Article

Figure 7. Plot of enzyme velocity as a function of inhibitor (I) concentration for chromone 1 (a) and chromone 2 (b). The solid curve drawn through the data points represents the best fit to the Morrison equation.

Figure 8. Plot of enzyme velocity as a function of inhibitor (I) concentration for N-(3-bromophenyl)-4-oxo-4H-chromene-3-carboxamide (a), chromone 3 (b), and chromone 4 (c). The solid curve drawn through the data points represents the best fit to the Morrison equation.

Figure 9. Cellular inhibition of ROS production by inhibitors 1 (left panel) and 2 (right panel). HEK-293 cells transfected with MAO-B were incubated with tyramine for 15 min, and intracellular ROS generation was measured with the fluorescent probe DCFDA. Preincubation with Ldeprenyl (Dep, 1 μM) or compound 1 or 2 was performed for 10 min before addition of tyramine. Results are the mean ± SEM of four experiments: (∗∗∗) p < 0.001 for Ct vs tyramine; (∗) p < 0.01, (∗∗) p < 0.001 for tyramine vs tyramine + inhibitors, ANOVA post-Tukey’s statistical analysis.

affinity, but we cannot rule out the hypothesis that the lower rigidity of chromone 4 (due to the absence of the planar amide spacer) may destabilize inhibitor binding in the active site. Indeed, previous studies showed that analogs bearing a spacer with the carbonyl group but with either a sulfur or an oxygen atom in place of the NH group (i.e., lacking the full planar amide conformation and therefore featuring a lower rigidity) displayed higher IC50 values.13 Nevertheless, those analogs are expected to lack the intramolecular hydrogen bond that instead in 4 is likely to exist and confer a good level of rigidity to the molecule. Therefore, the hydrogen bond between chromone analogs and Cys172 is a key (though not unique) element in determining the high binding affinity of 1, 2, and 3 inhibitors for human MAO-B active site. Biological Evaluation of ROS Production Assay. On the basis of their potent effect in inhibiting recombinant human

binding inhibition profile could still be observed, chromone 4 acted as a competitive inhibitor with a Ki of 2927 ± 488 nM, i.e., about 100-fold higher than that of the other chromone analogs (Figure 8c). This suggests that although the chromone scaffold is likely to trigger the tight-binding inhibition mechanism of these compounds, the amide linker is essential for nanomolar binding affinity. Possibly, the hydrogen bond between the amide carbonyl group and Cys172 observed in the crystal structures (Figure 3b−d) may play a relevant role. To confirm this hypothesis, we attempted to determine the structure of human MAO-B in complex with chromone 4. Although X-ray data could be collected at 2.0 Å resolution, the electron density map for the inhibitor was of very bad quality, which hampered modeling of the molecule in the enzyme active site. Indeed, this supports the idea that the hydrogen bond with Cys172 mostly contributes to significantly increase binding 4209

DOI: 10.1021/acs.jmedchem.8b00357 J. Med. Chem. 2018, 61, 4203−4212

Journal of Medicinal Chemistry

Article

inhibitor L-deprenyl. In conclusion, chromone represents a valuable scaffold for the development of novel potent and reversible inhibitors of human MAO-B.

MAO-B, we next carried out a biological evaluation of chromones 1 and 2 on ROS production in HEK-293 cells transfected with human MAO-B and stimulated with the substrate tyramine. As shown in Figure 9, incubation of MAOB-transfected cells with tyramine for 15 min induced ROS production as measured by the 2′,7′-dichlorofluorescin diacetate (DCFDA) fluorescence assay. The specific involvement of MAO-B in ROS generation was supported by the complete prevention of DFCDA production by the MAO-B inhibitor L-deprenyl. As observed with L-deprenyl, preincubation of the cells with compound 1 or 2 dose-dependently inhibited ROS production following tyramine addition (Figure 9). Remarkably, both 1 and 2 were 1000-fold more potent than L-deprenyl for inhibition of ROS production by MAO B in HEK-293 cells. Pan Assay Interference Compounds (PAINS) Evaluation. Compounds were theoretically analyzed by means of ZINC PAINS Pattern Identifier23 and FAF-Drug4,24 and none were highlighted as PAINS or aggregator. Taking into account theoretical results and the experimental data, the compounds cannot be considered as false positive inhibitors.

■

EXPERIMENTAL SECTION

Reagents. Chromone-3-carboxylic acid, 3-formylchromone, sodium sulfate, dimethylformamide (DMF), phosphorus oxychloride (POCl3), aniline derivatives were purchased from Sigma-Aldrich. All other reagents and solvents were pro analysis grade and were acquired from Carlo Erba Reagents and Scharlab and were used without additional purification. All reagents for protein purification and crystallization and for enzymatic assays were purchased from SigmaAldrich except for detergents that were from Anatrace (USA). General Procedures. Thin-layer chromatography (TLC) was carried out on precoated silica gel 60 F254 (Merck) with layer thickness of 0.2 mm. For analytical control the following systems were used: ethyl acetate/methanol and chloroform/methanol in several proportions. The spots were visualized under UV detection (254 and 366 nm). Flash chromatography was performed using silica gel 60 0.2−0.5 or 0.040−0.063 mm (Carlo Erba Reagents). Following the workup, the organic phases were dried over Na2SO4. Solutions were decolorized with activated charcoal when necessary. The recrystallization solvents were ethyl acetate, dichloromethane, or ethyl ether/nhexane. Solvents were evaporated in a Buchi rotavapor. The purity of the final products (>97% purity) was verified by high-performance liquid chromatography (HPLC) equipped with a UV detector. Chromatograms were obtained in an HPLC/DAD system, a Jasco instrument (pumps model 880-PU and solvent mixing model 880-30, Tokyo, Japan), equipped with a commercially prepacked Nucleosil RP18 analytical column (250 mm × 4.6 mm, 5 μm, Macherey-Nagel, Duren, Germany) and UV detection (Jasco model 875-UV) at the maximum wavelength of 254 nm. The mobile phase consisted of a methanol/water or acetonitrile/water (gradient mode, room temperature) at a flow rate of 1 mL/min. The chromatographic data were processed in a Compaq computer, fitted with CSW 1.7 software (DataApex, Czech Republic). NMR data were acquired at room temperature on a Brüker AMX 400 spectrometer operating at 400.15 MHz for 1H and 100.62 MHz for 13C. Chemical shifts were expressed in δ (ppm) values relative to tetramethylsilane (TMS) as internal reference; coupling constants (J) were given in Hz. Assignments were also made from DEPT (distortionless enhancement by polarization transfer) (underlined values). Mass spectra (MS) were carried out on a Bruker Microtof (ESI) apparatus; the data were reported as m/z (% of relative intensity of the most important fragments). Synthesis of Chromone Carboxamides (1−3). To a solution of chromone-3-carboxylic acid (2.6 mmol) in DMF (4 mL), POCl3 (2.6 mmol) was added. The mixture was stirred at room temperature for 30 min for the in situ formation of the acyl chloride. Then, an aromatic amine with the desired aromatic pattern was added to the reaction. After 1−5 h, the mixture was diluted with dichloromethane (20 mL), washed with H2O (2 × 10 mL) and with saturated NaHCO3 solution (2 × 10 mL). The organic phase was dried, filtered, and concentrated under reduced pressure. The crude product was purified by flash chromatography and/or crystallization. N-(3′-Fluorophenyl)-4-oxo-4H-chromene-3-carboxamide (3). The compound was obtained in 40% yield and recrystallized from CH2Cl2. 1H NMR (400 MHz, CDCl3) δ 11.51 (1H, s, NH), 9.07 (1H, s, H2), 8.34 (1H, dd, J = 8.0, 1.3 Hz, H5), 7.80 (1H, dd, J = 2.0, 2.0 Hz, H2’), 7.72 (1H, ddd, J = 8.7, 7.1, 1.7 Hz, H7), 7.63−7.52 (2H, m, H8, H6), 7.39−7.24 (2H, m, H6′, H5′), 6.84 (1H, ddd, J = 8.0, 2.1, 0.9 Hz, H4′). 13C NMR (101 MHz, CDCl3) δ 177.4 C4, 164.2 C3′, 163.0 C2, 160.9 CONH, 156.2 C8a, 139.4 C1′, 135.0 C7, 130.1 C5′, 126.6 C5, 126.3 C6, 124.0 C4a, 118.6 C6′, 115.9 C2′, 115.8 C8, 111.3 C4′, 111.0 C3. ESI/MS m/z (%): 306 ([M + Na]+, 100). 3-(((3-Chlorophenyl)amino)methyl)-4H-chromen-4-one (4). To a solution of 3-formylchromone (250 mg, 1.4 mmol) in dichloromethane (5 mL), 3-chloroaniline (154 mg, 1.4 mmol) and Na(AcO)3BH (426 mg, 2 mmol) were added under inert atmosphere at room temperature. After 15 min, the mixture was diluted with

■

CONCLUSION A series of chromone 3-phenylcarboxamide analogs were previously identified as novel potent and reversible inhibitors of human MAO-B (with essentially no activity on MAO-A at 100 μM).13 As their exact mechanism of enzyme inhibition was not clarified and no experimental structural information was available, we carried out a crystallographic and biochemical analysis of a selected series of chromone analogs (Figure 2). Our studies showed that chromone inhibitors bind in the enzyme active site cavity with the 1,4-benzopyrone moiety ring in front of the flavin cofactor and they establish two hydrogen bonds with Tyr435 and Cys172 (Figure 3b−d). The position of the substituted exocyclic ring is also well conserved for all three chromone analogs, with the halogen atoms of 2 and 3 overlapping with one of the two methyl groups of 1 and with the fluorine atom of safinamide (Figure 4). The planar conformation of these inhibitors, enhanced by an intramolecular hydrogen bond between the pyrone carbonyl group and the amide spacer, perfectly matches the flat hydrophobic active site cavity of human MAO-B. These results are in agreement with most of the previous docking predictions11,14 and were interpreted in light of the tightbinding inhibition mode of these chromone analogs. As a matter of fact, the steady-state inhibition studies resulting in either mixed or noncompetitive inhibition profiles (confirmed also with the purified recombinant enzyme) were in contrast with the crystal structures showing all inhibitors bound in the substrate-binding site. Accordingly, a dose−response curve approach21 was adopted, which clearly revealed that chromone analogs are tight-binding inhibitors with Ki values of 55, 17, and 31 nM for 1, 2, and 3, respectively. Interestingly, we could still observe a tight-binding inhibition profile for chromone 4, which nevertheless showed a much lower binding affinity (Ki of 2927 nM). Although it is known that the chromone moiety itself displays a high affinity for MAO-B active site,11 we postulate that the hydrogen bond provided by the spacer carbonyl to Cys172 and the good fit of these analogs into the enzyme cavity significantly enhance the inhibitory activity. Remarkably, this was also observed in the cellular context with a 1000-fold stronger reduction of ROS levels in treated HEK-293 cells for both 1 and 2 compared to the well-known covalent MAO-B 4210

DOI: 10.1021/acs.jmedchem.8b00357 J. Med. Chem. 2018, 61, 4203−4212

Journal of Medicinal Chemistry

Article

Biological Evaluation of ROS Production. HEK-293 cells were grown in Dulbecco’s modified Eagle medium (DMEM) containing 10% heat-inactivated fetal bovine serum (FBS) under 5% CO2 at 37 °C. Cells were transfected with the plasmid pcDNA3.1-MAO-B (rat cDNA) using Lipofectamine (Thermo Fisher Scientific). Tyramine and L-deprenyl were from Sigma-Aldrich. Intracellular ROS were measured using the DCFDA probe assay (Thermo Fisher Scientific). Briefly, cells were loaded for 45 min with DCFDA probe (5 μM), washed with HBSS, and stimulated with tyramine (500 μM) in (Hanks’ balanced salt solution (HBSS) for 15 min, in the presence of the indicated inhibitors. Fluorescence was measured using Varioskan microplate reader (Thermo Fisher Scientific).

dichloromethane (40 mL) and washed with saturated NaHCO3 solution (2 × 10 mL). The organic phase was dried, filtered, and evaporated. The residue was purified by FC (dichloromethane) and recrystallized from dichloromethane. The compound was obtained in 60% yield. 1H NMR (400 MHz, CDCl3) δ 8.23 (1H, ddd, J = 8.0, 1.7, 0.4 Hz, H5)), 7.91 (1H, t, J = 1.0 Hz, H2), 7.67 (1H, ddd, J = 8.6, 7.1, 1.7 Hz, H7), 7.48−7.35 (2H, m, H6, H8), 7.06 (1H, dd, J = 8.0, 8.0 Hz, H5′), 6.68 (1H, ddd, J = 8.0, 2.1, 0.9 Hz, H4′), 6.63 (1H, dd, J = 2.1, 2.1 Hz, H2′), 6.52 (1H, ddd, J = 8.0, 2.1, 0.9 Hz, H6′), 4.24 (2H, d, J = 1.0 Hz, CH2). 13C NMR (101 MHz, CDCl3) δ 177.9 C4, 156.6 C8a, 153.1 C2, 148.6 C1′, 135.1 C3′, 133.8 C7, 130.3 C5, 125.7 C6, 125.3 C4′, 123.8 C4a, 120.8 C3, 118.2 C5′, 118.0 C8, 113.1 C2’, 111.8 C6′, 40.4 CH2. ESI/MS m/z (%): 285 ([M + 1]+, 20), 159 (96), 131 (100). Protein Purification and Enzymatic Assays. Human recombinant MAO-A and MAO-B were expressed in Pichia pastoris and purified as previously described.25,26 Purified protein samples were stored in 50 mM potassium phosphate buffer, pH 7.5, 0.8% (w/v) βoctylglucoside, 20% glycerol. Enzyme concentration was determined by measuring the absorbance UV/vis spectrum using a NanoDrop ND-1000 spectrophotometer (Thermo Scientific). MAO-A and MAOB activity was determined by a fluorescence-based method with benzylamine and kynuramine as substrates, respectively, which detects the increase in H2O2 (side product of the MAO reaction) with time using the horseradish peroxidase−Amplex Red coupled assay.20 The reaction was started by adding the enzyme sample at 0.7 nM final concentration in 50 mM Hepes buffer, pH 7.5, containing 0.25% (w/ v) reduced Triton X-100. All measurements were done at 25 °C using either a CLARIOstar plate reader (BMG Labtech) or a Cary Eclipse fluorimeter (Agilent) depending on the type of experiment. Steady-State Kinetic Analysis. The MAO-B inhibition mechanism of the chromone analogs was determined by measuring the initial rates of substrate oxidation in the presence of varying concentrations of inhibitor. Data were fit to the appropriate form of the Michaelis− Menten equation and analyzed using GraphPad Prism software 5.0. The mode of inhibition was determined using global fit analysis of the enzyme velocity versus substrate concentration curves in the presence and absence of the inhibitor to equations for competitive, mixed, noncompetitive, and uncompetitive inhibition. Visual inspection of the fitting curves and evaluation of the r2 value were used to select the best fitting for determination of the inhibition constants. The results were also analyzed as double reciprocal plots (1/kcat vs 1/[S]). Tight-Binding Inhibition Assays and Ki Determination. Further inhibition studies were carried out using the same fluorescence-based assay described above, following protocols developed for tight-binding inhibition.21 In particular, the tightbinding inhibition mechanism for each chromone analog was evaluated by measuring the IC50 values (obtained by fitting percentage of inhibition data versus inhibitor concentration using GraphPad software 5.0) at four fixed substrate concentrations and plotting these values versus substrate concentration. Next, Ki values were determined through a dose−response curve by plotting the enzyme activity values at a fixed substrate concentration (2.7 mM benzylamine) and at varying concentrations of each chromone inhibitor using the Morrison equation27 in GraphPad Prism software 5.0. X-ray Crystallography. Human MAO-B in 50 mM potassium phosphate, pH 7.5, 8.5 mM Zwittergent 3-12 was cocrystallized with the chromone inhibitors by the sitting-drop vapor diffusion method following published protocols.19 X-ray diffraction data were collected at the beamlines of the Swiss Light Source in Villigen (Switzerland) and European Synchrotron Radiation Facility in Grenoble (France). For data collection, crystals were transferred into a mother liquor solution containing 18% (v/v) glycerol and flash-cooled in a stream of gaseous nitrogen at 100 K. Data processing and scaling (Table 1) were performed using XDS28 and the CCP4 package.29 The coordinates of the MAO B-safinamide complex,15 deprived of all water and inhibitor atoms, were used as initial model. The program Coot30 was used for electron density inspection and model building, whereas crystallographic refinement was performed with the program REFMAC5.31 Figures were generated by the program CCP4mg.32

■

ASSOCIATED CONTENT

S Supporting Information *

The Supporting Information is available free of charge on the ACS Publications website at DOI: 10.1021/acs.jmedchem.8b00357. Molecular formula strings (CSV) Accession Codes

Atomic coordinates and experimental data of human MAO-B in complex with inhibitors 1, 2, and 3 were deposited in the Protein Data Bank with PDB codes 6FVZ, 6FW0, and 6FWC, respectively. Authors will release the atomic coordinates and experimental data upon article publication.

■

AUTHOR INFORMATION

Corresponding Author

*E-mail: [email protected]. Phone: (+39) 0382-985527. ORCID

Fernanda Borges: 0000-0003-1050-2402 Claudia Binda: 0000-0003-2038-9845 Notes

The authors declare no competing financial interest.

■

ACKNOWLEDGMENTS Fellowships for J.R. (Grant SFRH/BD/96033/2013) and F.C. (Grant SFRH/BPD/74491/2010) are supported by FCT and FEDER/COMPETE funds. N.M.’s postdoc fellowship was supported by Fondazione Cariplo (Grant 2014-0672 to C.B.). The COST action CA15135 (Multi-Target Paradigm for Innovative Ligand Identification in the Drug Discovery Process MuTaLig) and Società Italiana di Biofisica e Biologia Molecolare (SIBBM) are kindly acknowledged for supporting visiting fellowships to J.R. We thank the European Synchrotron Radiation Facility (ESRF) and the Swiss Light Source (SLS) for providing beamtime and assistance and the European Community’s Seventh Framework Programme (FP7/2007− 2013) under BioStruct-X (Grants 7551 and 10205) for funding synchrotron trips. J.M.-P. and A.P. were funded by Région Occitanie. This work was supported by Fondazione Cariplo (Grant 2014-0672 to C.B.), Foundation for Science and Technology (FCT), FEDER/COMPETE2020 (UID/QUI/ 00081/2015, and POCI-01-0145-FEDER-006980).

■

ABBREVIATIONS USED CNS, central nervous system; DCFDA, 2′,7′-dichlorofluorescin diacetate; DMEM, Dulbecco’s modified Eagle medium; DMF, N,N-dimethylformamide; FAD, flavin adenine dinucleotide; FBS, fetal bovine serum; FC, flash chromatography; MAO-A, monoamine oxidase A; MAO-B, monoamine oxidase B; MD, molecular dynamics; ROS, reactive oxygen species. 4211

DOI: 10.1021/acs.jmedchem.8b00357 J. Med. Chem. 2018, 61, 4203−4212

Journal of Medicinal Chemistry

■

Article

Structures and Catalytic Properties of Tyr435 Mutant Proteins. Biochemistry 2006, 45, 4775−4784. (19) Binda, C.; Hubálek, F.; Li, M.; Herzig, Y.; Sterling, J.; Edmondson, D. E.; Mattevi, A. Crystal Structures of Monoamine Oxidase B in Complex with Four Inhibitors of the N-Propargylaminoindan Class. J. Med. Chem. 2004, 47, 1767−1774. (20) Zhou, M.; Panchuk-Voloshina, N. A One-Step Fluorometric Method for the Continuous Measurement of Monoamine Oxidase Activity. Anal. Biochem. 1997, 253, 169−174. (21) Copeland, R. A. Tight Binding Inhibitors. In Enzymes: A Practical Introduction to Structure, Mechanism, and Data Analysis, 2nd ed.; Wiley-VCH: New York, 2000; pp 304−317. (22) Cha, S.; Agarwal, R. P.; Parks, R. E. Tight-binding inhibitors II: Non-Steady State Nature of Inhibition of Milk Xanthine Oxidase by Allopurinol and Alloxanthine and of Human Erythrocytic Adenosine Deaminase by Coformycin. Biochem. Pharmacol. 1975, 24, 2187−2197. (23) Sterling, T.; Irwin, J. J. ZINC 15-Ligand Discovery for Everyone. J. Chem. Inf. Model. 2015, 55, 2324−2337. http://zinc15.docking.org/ patterns/home. (24) Lagorce, D.; Sperandio, O.; Galons, H.; Miteva, M. A.; Villoutreix, B. O. FAF-Drugs2: Free ADME/Tox Filtering Tool to Assist Drug Discovery and Chemical Biology Projects. BMC Bioinf. 2008, 9, 396−405. (25) Li, M.; Hubálek, F.; Newton-Vinson, P.; Edmondson, D. E. High-Level Expression of Human Liver Monoamine Oxidase A in Pichia pastoris: Comparison with the Enzyme Expressed in Saccharomyces cerevisiae. Protein Expression Purif. 2002, 24, 152−162. (26) Newton-Vinson, P.; Hubalek, F.; Edmondson, D. E. High-Level Expression of Human Liver Monoamine Oxidase B in Pichia pastoris. Protein Expression Purif. 2000, 20, 334−345. (27) Henderson, P. J. F. A. Linear Equation that Describes the Steady-State Kinetics of Enzymes and Subcellular Particles Interacting with Tightly Bound Inhibitors. Biochem. J. 1972, 127, 321. (28) Kabsch, W. XDS. Acta Crystallogr., Sect. D: Biol. Crystallogr. 2010, 66, 125−132. (29) The CCP4 Suite: Programs for Protein Crystallography. Collaborative Computational Project, Number 4. Acta Crystallogr., Sect. D: Biol. Crystallogr. 1994, D50, 760−767, DOI: 10.1107/ S0907444994003112. (30) Emsley, P.; Lohkamp, B.; Scott, W. G.; Cowtan, K. Features and Development of Coot. Acta Crystallogr., Sect. D: Biol. Crystallogr. 2010, 66, 486−501. (31) Murshudov, G. N.; Skubak, P.; Lebedev, A. A.; Pannu, N. S.; Steiner, R. A.; Nicholls, R. A.; Winn, M. D.; Long, F.; Vagin, A. A. REFMAC5 for the Refinement of Macromolecular Crystal Structures. Acta Crystallogr., Sect. D: Biol. Crystallogr. 2011, 67, 355−367. (32) McNicholas, S.; Potterton, E.; Wilson, K. S.; Noble, M. E. M. Presenting Your Structures: the CCP4mg Molecular-Graphics Software. Acta Crystallogr., Sect. D: Biol. Crystallogr. 2011, 67, 386− 394.

REFERENCES

(1) Edmondson, D. E.; Binda, C.; Wang, J.; Upadhyay, A. K.; Mattevi, A. Molecular and Mechanistic Properties of the MembraneBound Mitochondrial Monoamine Oxidases. Biochemistry 2009, 48, 4220−4230. (2) Ramsay, R. R. Monoamine Oxidases: The Biochemistry of the Proteins as Targets in Medicinal Chemistry and Drug Discovery. Curr. Top. Med. Chem. 2012, 12, 2189−2209. (3) Blair, H. A.; Dhillon, S. Safinamide: A Review in Parkinson’s Disease. CNS Drugs 2017, 31, 169−176. (4) de la Fuente-Fernández, R.; Sossi, V.; Huang, Z.; Furtado, S.; Lu, J.-Q.; Calne, D. B.; Ruth, T. J.; Stoessl, A. J. Levodopa-Induced Changes in Synaptic Dopamine Levels Increase with Progression of Parkinson’s Disease: Implications for Dyskinesias. Brain 2004, 127, 2747−2754. (5) Maggiorani, D.; Manzella, N.; Edmondson, D. E.; Mattevi, A.; Parini, A.; Binda, C.; Mialet-Perez, J. Monoamine Oxidases, Oxidative Stress, and Altered Mitochondrial Dynamics in Cardiac Ageing. Oxid. Med. Cell. Longevity 2017, 2017, 3017947. (6) Wu, J. B.; Yin, L.; Shi, C.; Li, Q.; Duan, P.; Huang, J.-M.; Liu, C.; Wang, F.; Lewis, M.; Wang, Y.; Lin, T.-P.; Pan, C.-C.; Posadas, E. M.; Zhau, H. E.; Chung, L. W. K. MAOA-Dependent Activation of ShhIL6-RANKL Signaling Network Promotes Prostate Cancer Metastasis by Engaging Tumor-Stromal Cell Interactions. Cancer Cell 2017, 31, 368−382. (7) Czech, M. P. Macrophages Dispose of Catecholamines in Adipose Tissue. Nat. Med. 2017, 23, 1255. (8) Gaspar, A.; Matos, M. J.; Garrido, J.; Uriarte, E.; Borges, F. Chromone: A Valid Scaffold in Medicinal Chemistry. Chem. Rev. 2014, 114, 4960−4992. (9) Reis, J.; Gaspar, A.; Milhazes, N.; Borges, F. Chromone as a Privileged Scaffold in Drug Discovery: Recent Advances. J. Med. Chem. 2017, 60, 7941−7957. (10) Gaspar, A.; Reis, J.; Fonseca, A.; Milhazes, N.; Viña, D.; Uriarte, E.; Borges, F. Chromone 3-Phenylcarboxamides as Potent and Selective MAO-B Inhibitors. Bioorg. Med. Chem. Lett. 2011, 21, 707−709. (11) Gaspar, A.; Silva, T.; Yáñez, M.; Vina, D.; Orallo, F.; Ortuso, F.; Uriarte, E.; Alcaro, S.; Borges, F. Chromone, a Privileged Scaffold for the Development of Monoamine Oxidase Inhibitors. J. Med. Chem. 2011, 54, 5165−5173. (12) Cagide, F.; Silva, T.; Reis, J.; Gaspar, A.; Borges, F.; Gomes, L. R.; Low, J. N. Discovery of Two New Classes of Potent Monoamine Oxidase-B Inhibitors by Tricky Chemistry. Chem. Commun. 2015, 51, 2832−2835. (13) Reis, J.; Cagide, F.; Chavarria, D.; Silva, T.; Fernandes, C.; Gaspar, A.; Uriarte, E.; Remiao, F.; Alcaro, S.; Ortuso, F.; Borges, F. Discovery of New Chemical Entities for Old Targets: Insights on the Lead Optimization of Chromone-Based Monoamine Oxidase B (MAO-B) Inhibitors. J. Med. Chem. 2016, 59, 5879−5893. (14) Ferino, G.; Vilar, S.; Matos, M. J.; Uriarte, E.; Cadoni, E. Monoamine Oxidase Inhibitors: Ten Years of Docking Studies. Curr. Top. Med. Chem. 2012, 12, 2145−2162. (15) Binda, C.; Wang, J.; Pisani, L.; Caccia, C.; Carotti, A.; Salvati, P.; Edmondson, D. E.; Mattevi, A. Structures of Human Monoamine Oxidase B Complexes with Selective Noncovalent Inhibitors: Safinamide and Coumarin Analogs. J. Med. Chem. 2007, 50, 5848− 5852. (16) Mazmanian, K.; Sargsyan, K.; Grauffel, C.; Dudev, T.; Lim, C. Preferred Hydrogen-Bonding Partners of Cysteine: Implications for Regulating Cys Functions. J. Phys. Chem. B 2016, 120, 10288−10296. (17) Echols, N.; Moriarty, N. W.; Klei, H. E.; Afonine, P. V.; Bunkóczi, G.; Headd, J. J.; McCoy, A. J.; Oeffner, R. D.; Read, R. J.; Terwilliger, T. C.; Adams, P. D. Automating Crystallographic Structure Solution and Refinement of Protein−Ligand Complexes. Acta Crystallogr., Sect. D: Biol. Crystallogr. 2014, 70, 144−154. (18) Li, M.; Binda, C.; Mattevi, A.; Edmondson, D. E. Functional Role of the “Aromatic Cage” in Human Monoamine Oxidase B: 4212

DOI: 10.1021/acs.jmedchem.8b00357 J. Med. Chem. 2018, 61, 4203−4212